WICSA 2005:Paper Reflections
From WICSA Conference Wiki
These are my reflections so far of the paper presentations in the WICSA conferences.
As a practitioner in an evolving development organisation heading for an efficient and effective development organisation, I come here looking for lean and efficient techniques that helps me deal with software architecture.
What I have seen so far has mostly been proposals for methods or processes that add a significant amount of effort to the development process. That might be OK were it not for the vagueness as to what will be easier to do later in the development process.
Any good architect knows he/she is expected to make significant decisions, based on conflicting stakeholder needs and many times incomplete information to base the decisions. Most architects are smart enough to come up with complex framworks and processes to formalise the way they like to architect systems, however, it is finding the leanest process and method that gets most of the issues sorted in a fraction of the time that is hard.
I guess what I look for is agile techniques for architecting. Any body?
Response from Chris Corbell Nov. 8, 2005
This has been my general impression also - most papers have presented theoretical or empirical techniques which would not be useful for real-world agile software development. I think that similar to usability concerns in software, there are usability concerns in architectural models - is the model or technique actually usable by an architect in a real-world scenario? Those of us who have this impression should perhaps attempt to enumerate some usability principles that we could give back to the software architecture community, esp. the academic community.
Response from Jeff Garland Nov. 8, 2005
I agree there is a very academic nature to many of the presentations -- perhaps WICSA needs to have an explicit "Experience Reports" track that would be more practical. That said, there have already been several WICSA presentations that are very practically oriented. Have you been to any tutorials? Having attended T01 and presented T05 I can say that both were very practically oriented and compatible with agile development. BTW, I also agree with Chris's point that practioners need to help codify problems for researchers if we ever expect to see researchers produce directly applicable results.
Response from PerOlof Bengtsson Nov. 8, 2005 I like the idea of usability assumptions. I guess what I experience is that the forces upon the architects in our organisations is different in some ways than researchers often understand them. With exceptions, of course. When I was doing my PhD work I thought I had a reasonable understanding, but I realise now that I have been in industry for a few years, that I didn't really understand all the forces (and maybe even misunderstood some) that controls the work performed in commercial software development.
Response from Robert Nord Nov. 10, 2005 The vision of working sessions envisioned by the conference organizers was to faciliate this kind of exchange. I think Mary Shaw in her Stevens Leture also articulated the contributions that can occur by looking for the dissonance that results from what we say and what we do.
Could we use this model to help us think about what we say in research and looking for dissonance with what we do in practice? How can we (the conference organizers) do a better job? An experience reports track that Jeff suggested is one way of accomplishing this. I'd like to hear more about this and other ideas.
